The Artificial Intelligence Initiative Act would devote resources to education and research, and help coordinate AI adoption across government.
A bipartisan trio of lawmakers wants to pour billions of dollars into expanding the use of artificial intelligence across government and industry, and preparing the country’s workforce to use the tech.
Sens. Martin Heinrich, D-N.M., Rob Portman, R-Ohio, and Brian Schatz, D-Hawaii, on Tuesday proposed a bill that would create a national artificial intelligence strategy and invest some $2.2 billion in advancing the tech over the next five years. The Artificial Intelligence Initiative Act aims to help the U.S. stay ahead of global competitors like China that are vying to dominate the international AI market.
“China’s emergence in the AI space poses a grave danger to an ethical, global adoption of standards and uses of these technologies,” Heinrich said during a call with reporters. “We simply cannot allow this type of unethical use of technologies to proliferate around the globe. That’s what will happen if we don’t step up our efforts here domestically.”
In February, the Trump administration released a national artificial intelligence strategy that called on agencies to ramp up investments in AI research and explore other ways to advance the tech across society. However, the plan included few specific policy proposals and no additional funding to support those efforts, which drew criticism across the tech community.
While Heinrich credited the administration for making AI a national priority, he said the latest proposal would provide the resources and governmentwide coordination needed to “preserve [the country’s] technological edge.” The legislation is intended to mesh with the White House strategy, not negate it, he said.
The bill would create multiple high-profile organizations within government to develop a national plan for AI research and coordinate efforts to adopt the tech at federal agencies. The proposed National AI Coordination Office would lead the government’s efforts with help from two advisory committees, one made up of agency heads and the other of non-government tech experts.
The legislation would charge the National Institute of Standards and Technology with creating guidelines to assess the effectiveness of various algorithms and the quality of training data. NIST would receive $40 million per year to fund those efforts from 2020 to 2024.
During the same period, the National Science Foundation would receive $100 million per year to stand up five new centers promoting research and education related to AI. One would explicitly focus on educating children about the tech and another would specifically serve people of color. In addition, the foundation would continue promoting research to explore algorithmic bias, AI explainability and privacy, as well as the broader social implications of the tech.
Under the bill, the Energy Department would also receive an extra $1.5 billion over five years to expand its AI research efforts.
Even with the bill’s additional funding, the U.S. would still be spending less than China on advancing artificial intelligence. But like other tech leaders, Heinrich doesn’t see money as the ultimate indicator of which country will come out on top.
“We will never compete with China from a federal spending level,” Heinrich said, “but we still have time to make targeted investments in the foundation of American ingenuity to develop AI responsibly and preserve our technological edge.”